Who Gets Hired, and Where
Ooak Data, a Y Combinator S26 company founded in 2024, is hiring across Paris and San Francisco while a team of four (three co-founders and one ML engineer) processes proprietary workflow data from 20 companies into reinforcement-learning environments for frontier AI labs. The public job board lists twelve salaried roles with a six-figure median, but the live postings tell a sharper story about where the founders are placing their next bets.
The split is geographic, functional, and legal. San Francisco carries the commercial weight: Zero G Talent's board data shows a GTM AI Labs role at $120,000–$150,000 and Zero G Talent's board data puts an Account Executive for Data Partnerships at $100,000–$140,000. Both sit in the city where the Y Combinator directory lists the company as "5 employees based in San Francisco" and where buyer conversations happen. Paris carries the build weight: Zero G Talent's board data found Head of Engineering at €100,000–€130,000, Head of Operations (€80,000–€100,000), Chief of Staff (€80,000–€100,000), and an ML Engineer (€70,000–€100,000). Two commercial roles in the U.S., four technical and operational leadership roles in France.
The French entity, Ooak Data SAS, is the legal anchor: incorporated on September 9, 2024, registered at 60 rue François Ier in the 8th arrondissement (RCS Paris 932 867 963), and designated as the exclusive processor for any partner data originating in the European Union. The company's legal notice states plainly that EU data "is processed and stored exclusively within the European Union," which makes the Paris office the compliance and infrastructure backbone for every enterprise customer subject to GDPR. The Paris site carries the heavier technical headcount — Head of Engineering, Head of Operations, Chief of Staff, ML Engineer — because the data-governance tooling that keeps EU-sourced datasets inside EU borders is there. The Chief of Staff role, also Paris-based, sits next to the product core.
San Francisco hosts the two board-listed commercial roles. The Y Combinator directory still shows that figure. That figure predates the Paris hiring wave visible on the board, but it signals where the go-to-market motion lives. The GTM AI Labs title implies a function that blends sales engineering, solutions architecture, and product feedback loops for early design partners. The Account Executive role, explicitly tied to "Data Partnerships," owns the conversation with U.S.-based companies willing to contribute proprietary workflow data in exchange for tailored RL environments. Both roles benefit from proximity to the YC network and the concentration of AI labs in the Bay Area.
The board shows twelve salaried roles total. Y Combinator's Work at a Startup portal lists seven; Station F lists eight. The gap aligns with the YC company description noting hiring in product, engineering, operations, marketing, and finance. The careers page frames the mandate broadly: "Help build the data layer for AI agents" and "We are a small, focused team building the data infrastructure that frontier AI labs use to train and evaluate their agent models." That language — "data layer," "RL environments," "real-world AI performance" — signals that product and research hires must speak both ML systems and the messy reality of enterprise data pipelines.
For candidates, the team-weight question resolves simply: if you sell complex technical products to research labs, the San Francisco roles are the entry point. If you build the pipelines, environments, and operational scaffolding that turn proprietary data into reinforcement-learning-ready training sets, the Paris roles are where the founding team is investing leadership capital first. The Chief of Staff role, spanning strategy, operations, and hiring, may be the clearest window into how the organization intends to scale. The two offices coordinate constantly, but they optimize for different outputs. One builds the environment; the other fills it.
What the Offer Looks Like
Ooak Data is bootstrapped, revenue-generating, and hiring from leverage; Y Combinator's launch post reported revenue runs well into seven figures across 20 company data ecosystems, and the three co-founders plus one ML engineer have kept the team tiny by design. That constraint shapes the compensation model: base salaries sit at market rates for Paris and San Francisco, but equity is the real lever. The company states it directly: "hires are highly leveraged and equity is a meaningful part of compensation." Offers are structured as base plus meaningful early equity given the bootstrapped, early stage.
The clearest picture comes from the board's live listings, which reflect what the company is actually offering candidates today.
Paris: The Build Side
| Role | Base (EUR/year) | Equity Range | Experience |
|---|---|---|---|
| Head of Engineering | 100,000–130,000 | 0.10%–0.30% (YC) | 6+ years |
| ML Engineer | 70,000–100,000 | 0.10%–0.30% | 6+ years |
| Senior Data Engineer | 65,000–95,000 | 0.10%–0.30% | 3+ years |
| Senior Software Engineer | 55,000–80,000 | 0.10%–0.30% | 6+ years |
| Chief of Staff | 80,000–100,000 | 0.50%–1.00% | 6+ years |
| Head of Operations | 80,000–100,000 | — | — |
| Founding Marketing Lead | 50,000–80,000 | 0.10%–0.50% | 6+ years |
Y Combinator postings for the same Paris roles show identical base ranges and equity bands, confirming the board data is current. The Chief of Staff role carries the widest equity slice, half a percent to a full percent, reflecting its proximity to the founders and the "highly leveraged" hiring thesis. Head of Engineering tops the base range at €130k; the Founding Marketing Lead starts lower at €50k but reaches 0.50% equity.
San Francisco: The Sell Side
| Role | Base (USD/year) | Equity Range | Location |
|---|---|---|---|
| GTM AI Labs | 120,000–150,000 | — | San Francisco, CA |
| Account Executive — Data Partnerships | 100,000–140,000 | — | San Francisco, CA |
The two U.S. roles are commercial. Both sit above the Paris engineering bases in absolute dollars, tracking Bay Area market rates. Neither listing discloses an equity column, but the company's stated philosophy (equity as a meaningful component) applies globally. A Pingojo listing for a Senior Software Engineer in San Francisco at "$10,000–$15,000" (dated 2026-10-08) appears to be a monthly figure ($120k–$180k annualized), roughly consistent with the board's GTM band.
How Equity Works
Every Paris technical and leadership role publishes an explicit equity range: a tenth to three-tenths of a percent for individual-contributor engineering and data roles, up to half a percent for the Founding Marketing Lead, and half to a full percent for Chief of Staff. The board data does not show equity for the two San Francisco commercial roles or for Head of Operations, but the careers page says "meaningful early equity given the bootstrapped, early stage," implying grants are negotiated per role and candidate. With a team of four today (three founders + one ML engineer), a 0.10% grant represents a meaningful slice of a pre-Series A cap table. The company is YC S26, so the 409A is likely still low; early joiners capture the steepest appreciation if the "hundreds of data ecosystems" acquisition plan executes.
Contract and Benefits
The LinkedIn hiring announcement (2026-09-05) notes: "CDI or freelance, both work for us, as long as you are in Paris (9e)." CDI (Contrat à Durée Indéterminée) is the French permanent contract; it carries statutory protections, unemployment insurance, and employer-paid social charges (~45% on top of gross). Freelance status shifts those costs to the contractor but allows higher daily rates. The company describes work as "remote-friendly and globally distributed," yet the Paris roles explicitly require presence in the 9th arrondissement. San Francisco roles list no such constraint. Benefits beyond statutory French coverage (health, pension, transport) are not detailed in any public posting; at this stage, the benefits package is the equity upside and the autonomy the founders emphasize: "No micromanagement, no unnecessary process. You will have real impact from day one."
What the Numbers Tell You
The compensation structure reveals the company's priorities: Paris is the engineering and operations hub; San Francisco is the commercial hub. Equity bands are transparent for Paris roles, a rarity at this stage, which lets candidates model upside. The median board salary of $110k across 12 roles suggests the company is not lowballing; it is pricing to attract senior talent who can operate without process. For a candidate, the calculation is straightforward: take market base, bet on the equity multiple, and accept that "everything still to build" (per the founders' September 2026 LinkedIn post) means the job is the build.
Inside the Hiring Funnel
Ooak Data runs a five-stage funnel compressed for a five-person founding team: apply via ooakdata.com or the YC portal, founder screen, technical deep-dive, work session or paid trial, then offer with equity. Candidates apply through the site or its YC and LinkedIn pages; each résumé routes directly to the founders (no centralized recruiting function).
The founder screen (about 5 days) filters for communication clarity and alignment with the company's core premise: real-world data, not synthetic proxies, is the bottleneck for frontier AI agents. The founders listen for evidence that a candidate has transformed messy, permission-heavy datasets into actionable insights and can explain those insights to non-technical stakeholders.
The technical deep-dive (about 7 days) involves discussion and/or exercise on RL environments, agent evaluations, or multimodal data pipelines, matched to the candidate's area. The company's public writing doubles as a filter: "AI agents still fail on real work: the messy, multi-tool, permission-heavy workflows inside actual companies. You can't synthesize it either; real org charts, permissions, and cross-tool dependencies don't come out of a generator. You have to source it from real companies and make it safe. That's the hard part, and it's what we do."
The work session or paid trial (about 7 days) is common at YC-stage labs: a paid trial or working session on a real problem before either side commits. That step exists because the company cannot proxy for "can you operate in the gap between sandbox and production" with a whiteboard. The trial is the signal.
The offer stage (about 3 days) includes base plus such equity.
Disqualifiers are practical, not pedigree-based. Candidates who treat anonymization as a post-hoc step or default to synthetic data examples tend to stall at the technical deep-dive. Applicants who haven't wrestled with real org charts, permission models, and cross-tool dependencies rarely advance past the founder screen.
Who Lasts
The profile that succeeds here is shaped by the intersection the company occupies: an applied research lab that ships revenue-generating product, a bootstrapped Y Combinator company working with frontier labs, and a four-person team already processing data from 20 companies. The YC page states the problem plainly: the same challenge described earlier. Candidates who need clean benchmarks or curated datasets will not last. The work is retrieval, anonymization, and evaluation design over messy, multimodal corporate exhaust — Gmail, Slack, Notion, Jira, Drive, SharePoint — transformed into "digital twins" that preserve structural complexity while stripping PII.
That pipeline demands a specific blend. The ML engineer role (€70k–100k in Paris) and Head of Engineering (€100k–130k) sit alongside a Head of Operations (€80k–100k) and Chief of Staff (€80k–100k) — all in Paris — while GTM and Account Executive roles ($120k–150k and $100k–140k) sit in San Francisco. The split is deliberate: Paris carries the research and core engineering weight; San Francisco carries the lab partnerships and data-supply relationships. People who thrive here operate across that divide. Their research page says their work covers "evaluation methodology, dataset design, and the gap between benchmark performance and real-world capability." They cite "hidden technical debt" (boundary erosion, entanglement, hidden feedback loops, undeclared consumers) as the failure modes that don't appear in sandboxes. That is the daily vocabulary.
Equity is part of every offer; the company describes itself as bootstrapped and "very early," with seed funding only from YC S26. Cash compensation is real; board data shows a six-figure median across roles, but the bet is on the founding-era equity.
Remote-first global is the operating model, not a perk. The Paris cluster (engineering, ops, chief of staff) and SF cluster (GTM, partnerships) coordinate across time zones. The team is 1–10 people total. Everyone writes, everyone ships, everyone talks to customers or lab partners. The Chief Data Officer role they're actively recruiting signals the next phase: scaling the data-supply flywheel while the research team pushes evaluation methodology.
The founding-era window is the offer. A 0.10% equity grant on such a cap table with four people on it. The next hire will be a Chief Data Officer, the role they're already recruiting to scale the data-supply flywheel. The people who stay are the ones who treat that gap between benchmark and production as the product.
Working in AI? Zero G Talent tracks the openings: see every open Ooak Data role, browse AI jobs, the companies hiring, and the people building the field.